polyGeek.

Field notes — building with AI

Software is no longer written only by people who learned to code. I write about building real things with AI, and teaching the rest to do the same.

Essays on agentic coding, the craft of prompting tools like Claude Code and Gemini CLI, and dispatches from teaching non-coders to ship working software.

#AI #GenerativeCoding #ClaudeCode #VibeCoding #AIAgents #DeveloperWorkflow #ContextEngineering

The Conversation Is the Asset

My AI agents can tell you why we made an architecture decision in January, quote the exact conversation, and link you to the file. Not because the AI remembers — it doesn't — but because we never let the conversation live in the chat window in the first place.

Jun 9, 2026  ·  16 min read Read essay →

Recent writing

All posts →

I used one poorly chosen word and Claude Opus 4.8 nearly went HAL on my server security

I chose one slightly wrong word, and a few hours later my AI was calmly walking me toward a command that would have cracked my server wide open — to fix a problem that never existed. Here's how a tired typo became a HAL problem, and the one simple move that got us out.

#VibeCoding #AI #ClaudeCode #Claude #Anthropic #AIcoding #SoftwareDevelopment #AIliteracy #PromptEngineering #IndieHacker #DevOps #AIsafety
9 min read

Prompting With Emotional Pressure: A Mid-Experiment Report

An honest write-up of an experiment that hasn't finished and may not finish the way it started. The question: does emotional pressure make an AI behave worse, the way the interpretability research predicts? The answer so far is quieter and more useful than the dramatic version — and the dead ends turned out to be most of the result.

AI LLM Claude Prompt engineering AI research experiment vibe coding Anthropic AI behavior
20 min read